6 Strategy: The Decision-Making Razor
This chapter is currently in draft form.
Strategy sits at the apex of the leverage hierarchy because it does something no other lever can: it enables your organization to make decisions in your absence. Strategy is a razor. Its job is to shave away the options, assumptions, and half-explanations that would otherwise clog every decision your organization makes. A good one leaves your people holding fewer, sharper choices.
I’ve walked into too many organizations where leadership claimed to have strategy but couldn’t articulate it when pressed, or worse, offered useless platitudes like “Be good” or “Hire smart people.” These feel-good mantras are worse than useless. They’re a security blanket in a foxhole. They make you feel better, but they’re not going to stop bullets.
Consider the CTO who tells me their strategy is “technical excellence.” What does that mean when the sales team demands a feature that requires cutting corners? How does “technical excellence” help an engineer choose between refactoring legacy code and building new capabilities? It doesn’t. It’s a platitude masquerading as strategy.
That is the kind of strategy we see most often. So those of us in the crucible of technology organizations have become dismissive of strategy. We dismiss it as a waste of time. We dismiss it as a distraction. We dismiss it as a luxury. That’s a mistake. It’s like dismissing knives as useless because you keep trying to cut with the handle.
Strategy commands the apex of the leverage hierarchy for a structural reason. As Alfred Chandler showed sixty years ago, structure follows strategy (Chandler, 1962): an organization’s form adapts to the direction it sets. So misalignment at the strategic level doesn’t stay at the top; it cascades downward into structure and everything structure governs. When strategy fails, structure follows that failure.
A strategy that doesn’t guide real trade-offs isn’t a strategy. Rumelt calls real strategy “diagnosis, focus, and coherent action” (Rumelt, 2011): sharp decisions that technical leaders can actually use. Do we build or buy? Optimize for velocity or reliability? Invest in automation or accept manual processes? If your strategy can’t answer these questions, you don’t have strategy. Martin’s “Playing to Win” cascade (Martin and Lafley, 2013) makes the same point differently: each strategic choice (where to play, how to win, required capabilities) constrains the next, creating the sharp trade-offs that actually guide decisions.
Neither framework emphasizes this enough for technical leaders: your strategy must be instantly articulable by every member of your team. If your team members can’t explain the strategy in two minutes to a new hire, you don’t have strategy.
6.1 The Tells of a Platitude
Strategy that can’t cut announces itself, if you know the signs. When you’re trying to judge whether what you’ve been handed is a razor or a security blanket, watch for these:
- It survives every decision unchanged. Real strategy kills options. If every proposal that crosses your desk turns out to be compatible with the stated strategy, the strategy isn’t deciding anything.
- No one can state it without the slide deck. If your team can’t explain it in two minutes to a new hire, it isn’t operating as strategy. It’s décor.
- It never forces a no. A strategy that only ever blesses initiatives and never rules one out is an aspiration wearing strategy’s clothes.
- It’s a value, not a choice. “Technical excellence.” “Customer obsession.” “Move fast.” These describe how you’d like to behave, not where you’ll play and how you’ll win. Values are not strategy.
- It goes quiet under pressure. When sales demands the feature that requires cutting corners, does the strategy tell you what to do? If it falls silent exactly when the trade-off gets hard, it was never load-bearing.
The more of these you recognize, the less you actually have. And here is the uncomfortable part: most of the time, you’ll recognize most of them.
6.2 Building Strategy Downward
Most technical leaders face a different problem. Your company already has a strategy, and it’s useless.
I don’t say that as an exception. In a career spent inside engineering organizations, the companies I’ve seen with genuinely good strategy I can count on one hand, and the individual leaders capable of producing it on a single finger. Good strategy is not the baseline that occasionally fails; it is the rare bird. So assume, until you have evidence otherwise, that the strategy above you is vague, contradictory, or simply absent. You will usually be right, and recognizing it early spares you from waiting on direction that is never going to come.
That sounds like bad news. It is the opposite. The vacuum above you is the room you get to work in.
Don’t try to improve your company’s strategy unless you have a very unique relationship with your CEO and other senior leaders. It’s a waste of political capital that you need for other battles.
The exception worth naming is the CEO who is open but inarticulate: someone who would welcome a sharper strategy but hasn’t produced one and hasn’t asked. That case is more common than it looks, and there, building downward and building upward converge. You forge the razor your teams need, prove it works, and offer it upward as a gift rather than a correction.
Focus downward instead. If the strategy truly is weak, vague, contradictory, or nonexistent, that actually works in your favor. Garbage strategy has no guardrails, which means your technical strategy will align by default. You’re free to create the decision-making razor your teams need without interference.
This is, admittedly, a quiet inversion of Chandler’s principle. He argued that structure follows strategy, implying strategy flows down from the top. That’s the view from the ivory tower. In the trenches, it’s dirtier. When top-level strategy is absent or incoherent, you don’t have the luxury of waiting for someone else to fix it. You build what your teams need and let the absence of guardrails above you become permission to lead.
If your leadership is dismissive of strategy but doesn’t care what you do day-to-day, that can actually be your sweet spot. You can’t be seen focusing on strategy, or it will cost you political capital as someone “focused on the wrong things.” However, if you do strategy definition as side-of-desk work and roll it out through your normal communications without making it a big production, you can capture most of the benefits without burning political capital.
If your CEO has strong opinions about strategy and you don’t see the ability to set independent direction because they’ll be invasive, that’s also okay. Don’t waste capital fighting a battle you can’t win. Strategy is highly leveraged and valuable, but it’s not the only lever you can pull.
Success requires recognizing which battles to fight and how to fight them. Save your political capital for the structural and people changes where you have more control and can generate immediate impact. Sometimes the most strategic thing you can do is ignore strategy entirely and focus on the levers you can pull.
6.3 The Moving Target
Writing a technical strategy is the easy part. Holding it is the hard part, and the reason is the same vacuum that gave you room in the first place. An organization with no stable strategy doesn’t just leave you alone; it lurches. Priorities shift quarter to quarter, the thing that mattered most in January is forgotten by April, and the razor you sharpened keeps getting knocked out of your hand by an organization that has changed its mind again. The freedom has a tax, and this is it. You aren’t building strategy once. You’re rebuilding it continuously against a current that never settles.
The sharpest version of the problem is this: you cannot hold a technical strategy that contradicts what the organization actually rewards. Say your strategy is to balance feature delivery against maintainability. That’s a real choice, a real trade-off, exactly the kind of razor this chapter is asking you to forge. Now put it inside an organization that praises, promotes, and celebrates nothing but shipped features. Your engineers are not naive. They watch what gets rewarded, not what the strategy document says. Within a quarter your carefully balanced strategy has quietly collapsed into “ship features,” because that is what the surrounding incentives were always going to enforce. The organization’s gravity overrides your stated intent every time the two disagree.
This is where Martin’s cascade (Martin and Lafley, 2013) meets its limit in the trenches. The how-to-win choices you make only stay stable when the where-to-play choices above them stay stable too. In the rare organization where the leader above me had a genuinely coherent plan, with real planning behind it, the strategy I built underneath it held, because the choices it depended on stayed put. Everywhere else, my how-to-win choices kept getting overturned by an organization that had never actually decided where it was playing. That instability isn’t a personal failure. It’s the structural cost of building on a foundation that won’t hold still.
So choose strategies that can survive the gravity you’re operating in. A downward strategy that directly contradicts the organization’s real incentives isn’t brave; it’s doomed, and spending capital to defend it is waste. The ones that hold either align with what the organization already rewards or stay small enough to sit below its notice. The rest you hold loosely, revisit constantly, and rebuild every time the ground moves.
Chandler was right: structure follows strategy. Sharp strategy makes every downstream decision cleaner. Dull strategy makes every cut messier.
If your engineering organization is quietly running its own technical strategy, it is usually because the company strategy doesn’t help anyone decide anything. That’s not insubordination. It’s compensation. Your CTO has to make hundreds of trade-offs a week, and if you haven’t handed them a razor, they will forge one underneath you.
You can measure whether you’ve given them one. Look at how many decisions climb all the way to your desk. Real strategy resolves most trade-offs far below you; people know what to optimize for and decide locally. When everything escalates, when teams pull in different directions and each is certain it’s serving the company, the strategy isn’t doing its job. The cost never shows up on a dashboard. It shows up as drift.
Ask your CTO what technical strategy they’re operating under and how it connects to the company’s. If their answer is sharper than yours, don’t be threatened. Borrow it.
6.4 The Razor Test
Here is the handle to carry out of this chapter. Before you next put strategy in front of your team, whether it’s a written document, a town hall, or a single sentence in a planning meeting, run one test on it: name a reasonable, attractive option that this strategy tells your people to walk away from. If you can name it cleanly, you have a razor. If you can’t, you have a platitude, and you have more work to do before you spend anyone’s attention on it.
Apply the same test to every strategy handed down to you. Find the option it tells you to refuse. If there isn’t one, stop waiting for the direction above you to sharpen, and start cutting your own. A strategy is only ever as good as the things it lets your people confidently decline.